The therapeutic potential of mesenchymal stem cells (MSCs) was discovered decades ago. However, despite extensive preclinical and clinical studies during such a long period of time, the molecular and cellular mechanisms underlying the beneficial effects of MSC transplantation have not been fully elucidated. Initially, it was assumed that transplanted MSCs exert therapeutic activity mainly through homing into damaged tissues [1,2]^[1][2] with subsequent differentiation into specialized types of targeted tissue cells (site-specific differentiation) replacing the dead or damaged cells [3,4,5]^[3][4][5]. Later, this mechanism of action was questioned since the site-specific differentiation of transplanted MSCs in the recipient’s tissues was shown to be an exceptionally rare event ^[6]. At the next stage of research targeted at the mechanisms of MSC action, it was found that after homing to the injury site, the transplanted MSCs are able to recruit resident stem/progenitor cells, which ensure subsequent tissue regeneration [7,8]^[7][8]. This recruiting ability was associated with the paracrine effects of MSCs, which produce a large number of trophic factors including cytokines, chemokines, and growth factors. In addition to resident progenitor cells recruiting to the damaged area, the paracrine effects of MSCs were also associated with their immunomodulatory abilities, the suppression of apoptosis of the resident cells in injured tissues, stimulation of angiogenesis, and enhancement of other regeneration-related processes [9,10]^[9][10]. By this time, some studies have reported a lack of the migration of transplanted cells and their rapid death [11,12]^[11][12]. Based on these discoveries, it was suggested that the transplantation of MSCs themselves is not necessary for the implementation of their regenerative potential, and, perhaps, conditioned media from MSCs could exhibit similar effects. Indeed, a number of studies have shown that MSC-conditioned media really enhance regeneration and mimic other MSC effects in various disease models [13,14,15]^[13][14][15]. Consequently, a new direction of biomedical research concentrated on the use of MSC-derived exosomes as therapeutic agents started to develop ^[16]. At the same time, data reliably demonstrating that transplanted MSCs undergo rapid clearance in the body of the recipient animals and do not have the ability for any long-term and effective homing have been steadily accumulating [11,17]^[11][17]. It was also found that apoptotic MSCs exhibit therapeutic effects, in particular, a pronounced immunomodulation activity ^[18]. Currently, an increasing number of researchers are inclined to believe that the apoptosis of transplanted MSCs is necessary for the manifestation of their regenerative and therapeutic potential. Accordingly, the administration of MSC-ApoBDs instead of whole live cells is being tested. However, apoptosis of MSCs after transplantation is still generally considered a disadvantage or as an event that negatively affects therapeutic efficacy, and researchers seek to increase the survival of transplanted cells in vivo [19,20,21]^[19][20][21]. Probably, the two opposite approaches can be reconciled based on the accumulating results of preclinical and clinical studies demonstrating that live MSCs and apoptotic MSCs are not therapeutically interchangeable, but rather are effective in different contexts and probably through different mechanisms ^[22].

2. Biodistribution of Transplanted MSCs Undergoing Apoptosis In Vivo

As mentioned above, in the last few years, more and more works have appeared where it has been reliably proven that transplanted MSCs, regardless of the method of their administration, are quickly eliminated, while the therapeutic effects are preserved. These results led to the idea that transplanted cells do not need to remain viable for a long time in order to realize their therapeutic effects. Moreover, probably not live, but apoptotic cells are responsible for at least the immunomodulatory effects of MSCs transplantation. As a rule, the bulk of the transplanted human MSCs isolated from bone marrow, adipose tissue, or the umbilical cord was found in the lungs of recipient animals within a few minutes after intravenous (i.v.) transplantation. Later, at 2 h after i.v. injection tiny numbers (less than 0.1%) of human MSCs were found in the liver. Transplanted human MSCs showed signs of apoptosis, including a decrease in cell size, caspase 3/7 expression, positive staining with annexin V, and co-expression of calreticulin, which is an “eat me” signal for phagocytic cells, as early as 30–60 min after the intravenous injection [118,119,120]^[23][24][25]. The number of living human MSCs progressively decreased in animals over time. All authors agree that human MSCs were completely eliminated and were not detected in the organs of recipient animals 24 h after transplantation [118,119,120]^[23][24][25]. Importantly, transplanted human MSCs that have undergone apoptotic cell death in animal bodies, i.e., in a xenogeneic environment, were co-stained with antibodies against the macrophage marker F4/80, the complement component C3b, the granulocyte marker GR-1, and the endothelial and platelet marker CD31/PECAM-1. These results indicate an interaction between human MSCs and different subpopulations of host phagocytic cells, which can potentially provide their rapid clearance [118]^[23]. As shown in several studies, the complete clearance of intravenously transplanted syngeneic MSCs was somewhat slower than in xenogeneic transplantation. The number of cells remaining alive decreased markedly after 12–24 h [121]^[26], and live cells were not detected after 7 days [122]^[27]. Similar results were obtained with intrapancreatic and intrasplenic transplantation of syngeneic MSCs when the complete clearance of transplanted cells occurred within 7 days. Moreover, transplanted cells were found in the liver 12 h after intrasplenic infusion, followed by a fall after 72 h [123]^[28]. In all the works described above, a significant increase in the expression of caspase 3 was noted in transplanted cells [121,122,123]^[26][27][28]. Subcutaneously injected syngeneic MSCs were detected a little longer (up to 7 days) with complete clearance after 14 days [122]^[27]. In the rat model of the brain injury, it was shown that intra-arterial infused human placenta MSCs [124]^[29] or human BM-MSCs [125]^[30] were not detected in rat brains 72 h after infusion. Transplanted cells were localized in the lumen of the blood vessels being in close contact with the vascular wall at 24–48 h after administration. At 72 h after transplantation, MSCs were phagocytosed, presumably, by activated microglia and macrophages, as shown by co-staining with CD44 and ED1 [125]^[30]. Thus, it has been shown that transplanted MSCs rapidly undergo apoptotic cell death followed by clearance due to host phagocytic cells, regardless of the route of administration, the source of MSCs, and even the genetic compatibility of the transplanted cells and the recipient animals. This raises several very important questions. First, what are the mechanisms of apoptosis induction in transplanted MSCs, and what factors trigger the process? Second, what phagocytic cells are able to ‘efferocytize’ apoptotic MSCs, and what are the consequences of this engulfing?

3. Apoptosis of Transplanted MSCs In Vivo

At present, very little is known about the mechanisms and molecular signals involved in triggering the cell death of transplanted MSCs. An important result of research in this area is the fact that apoptosis of transplanted MSCs is not the result of allogeneic and/or xenogeneic cell recognition since efficient elimination of infused cells occurs in syngeneic models as well. Galleu et al. [126]^[31] demonstrated that MSC apoptosis after transplantation was associated with the presence of effector cytotoxic CD8⁺Vβ8.3⁺ cells in the lungs of the GvHD recipient mice. Activated cytotoxic CD56⁺ NK cells and CD8⁺ T cells were found to be responsible for initiating MSC apoptosis. The molecular inducers of apoptosis included granzyme B and perforin and, to a lesser extent, FasL, while TRAIL was not involved in triggering the process [126]^[31]. The involvement of FasL/Fas in the induction of apoptosis of implanted MSCs was shown in a model of myocardial ischemia ^[19]. On the other hand, Pang et al. [120]^[25] tracked the presence of luciferase-expressing human MSCs in immunocompetent BALB/c mice and the two immunodeficient mouse models, NOD/SCID/Il2rγc^−/− (NSG) and BALB/c NOD.sirpa Rag2^−/−Il2rγc^−/− (BRGS). The authors concluded that the rapid clearance of MSCs in the absence of the adaptive immune response and cytotoxic cells suggests that MSCs die in the lungs due to factors unrelated to immune cell function (for example, in this case, cell death was most likely triggered by nutrient deprivation/growth factors), and clearance occurs by myeloid cells without regard for the state of inflammation. Preda et al. [122]^[27] suggested that apoptosis of transplanted MSCs may be associated with microenvironmental conditions, including hypoxia and pro-inflammatory cytokines such as TNFα and IFNγ. In earlier studies, it was shown that human BM-MSCs, despite the expression of three cell-surface complement regulators (CD46, CD55, and CD59), activated the complement system and, as a result, were injured by membrane attack complexes, which led to the lysis of transplanted cells after transplantation into immunocompetent mice [127,128]^[32][33]. Based on the available data, it is still difficult to identify molecular factors inducing the apoptotic death of transplanted MSCs. Apoptotic stimuli likely depend on the localization of the infused cells (lungs, liver, brain, other organs), on the microenvironment (hypoxia, inflammation, necrosis, etc.), and also on the presence of different types of immune cells at the site of the transplanted cell localization.

4. Transplanted Apoptotic MSCs Clearance and the Mechanisms of Their Therapeutic Action

Phagocytosis is a key regulator of tissue homeostasis and cell turnover in adulthood and development. It is known that not only ApoBDs, but some apoptotic cells expressing appropriate signals, including find-me signals, eat-me signals, don’t-eat-me signals, and opsonins, do not break down into ApoBDs, but are efficiently ‘efferocyted’ by phagocytes [129,130]^[34][35]. This section addresses the mechanisms underlying the clearance of transplanted MSCs undergoing apoptosis in vivo, clearance of transplanted apoptotic MSCs, and clearance of transplanted MSC-derived ApoBDs with regard to their therapeutic activity. In an elegant study, Luk et al. [131]^[36] showed that heat-inactivated MSCs that lost the capacity to respond to inflammatory stimuli and the ability to secrete trophic factors can still modulate the immune responses to sepsis, suggesting that MSCs can act as passive immunomodulatory vehicles. These inactivated MSCs did not suppress T cell proliferation and did not induce the formation of regulatory B cells, but modulated the function of monocytes in vitro. The authors concluded that the immunomodulatory effects of inactivated MSCs in vivo were associated precisely with their effect on the recipient’s monocytes [131]^[36]. A series of earlier studies conducted on animal models of the sepsis syndrome showed that apoptotic AD-MSCs exhibit protective effects associated with suppression of inflammation, inhibition of oxidative stress, and reduction of apoptosis in damaged organs [132,133,134,135]^{[37][38][39][40]}. Unfortunately, the papers described above there contain no data on the clearance of transplanted cells and the putative mechanisms underlying the therapeutic effects. In addition to the rapid death of transplanted MSCs, it has also been proven that the apoptotic cells effectively undergo clearance in the recipient’s body. This process takes no more than a few hours, depending on the route of cell administration. Here the determination of cells able to rapidly phagocytize/efferocytize apoptotic MSCs in vivo is very important since in this case, phagocytic cells are a kind of mediator between infused MSCs, and the implementation of their therapeutic effects. Several studies have shown a pronounced immunosuppressive/immunomodulatory effect of transplanted apoptotic MSCs and MSC-ApoBDs in animal models of various diseases, including allergic airway inflammation [120^[25][41],136], type 2 diabetes [113]^[42], liver fibrosis [123]^[28], acute liver injury, LPS-induced lung injury, and spinal cord injury [121]^[26], acute colitis [137]^[43]. In all the above works, apoptotic MSCs or MSC-ApoBDs were subject to efferocytosis mainly by macrophages. Thus, in a type 2 diabetes (T2D) mouse model, i.v. infused human BM-MSC ApoBDs showed marked accumulation in the liver after 24 h, where they were mainly phagocytosed by Kupffer cells and monocyte-derived macrophages. Presumably, the “eat me” signal for efferocytosis, in this case, was calreticulin exposed on the ApoBD surface. Efferocytosis of MSC-ApoBDs resulted in transcriptional reprogramming of macrophages by proteins contained in the ApoBDs that have the potential to induce macrophage polarization into an anti-inflammatory M2 phenotype, including alpha-crystallin B chain (CRYAB), cAMP-dependent protein kinase type II-alpha regulatory subunit (PRKAR2A), receptor of activated protein C kinase 1 (RACK1) and vasodilator-stimulated phosphoprotein (VASP). Thus, efferocytosis of MSC-ApoBDs inhibited diet-induced obesity-induced macrophage activation in the liver, leading to an improvement in the chronic inflammatory environment in the T2D model [113]^[42]. Apoptotic MSCs pretreated with staurosporine exerted immunosuppressive effects in the lungs and inhibited allergic asthma to the same extent as live MSCs. When live CTV-labelled MSCs were intravenously administered to BALB/c mice, the CTV label was detected in both CD45^- stromal and CD45⁺ hematopoietic subpopulations in the lung. Tracking CTV⁺CD45⁺ cells over time showed that there is a hierarchy of phagocytic cell types that engulf CTV-labelled MSCs in the lungs. Ly6G⁺ neutrophils were the predominant cell type phagocytizing MSCs 10 min after i.v. injection, Ly6Chi, and Ly6Clo monocytes mainly engulfed transplanted MSCs after 1 h, CD11b⁻CD103⁺ type 1 conventional dendritic cells (cDC1)—after 2 h, and CD64+ interstitial macrophages—after 4 h. Interstitial macrophages and cDC1 were the main phagocytic cells at 8 h after i.v. injection of MSCs. During the first 8 h, uptake of CTV-labeled MSCs by CD11c⁺SiglecF⁺ alveolar macrophages remained constant at a low level, while uptake by CD11b⁺ cDC2 remained consistently low. Alveolar macrophages were shown to be a critical resident phagocytic population of the lungs, clearing both live and apoptotic MSCs, which underwent phenotypic changes after efferocytosis. Efferocytosis of apoptotic MSCs induced IFN-responsive genes and metabolic reprogramming of alveolar macrophages, followed by a marked inhibition of lung inflammation. Taken together, these data demonstrate that efferocytosis of apoptotic MSCs induces sustained changes in the immunometabolism and alveolar macrophage function that directly inhibit lung inflammation. The symptoms of the disease were suppressed even several weeks after the elimination of MSCs from the lungs [120]^[25]. The protective effect of the dead MSCs which undergo spontaneous cell death during in vitro culturing was also confirmed in four mouse models, including concanavalin A (ConA)- and carbon tetrachloride (CCl₄)-induced acute liver injury, LPS-induced lung injury, and spinal cord injury. In the model of acute liver failure, i.v. transplantation of dead MSCs resulted in a significant decrease in activated NK cells and infiltrating neutrophils in the liver, while the number of Ly6Chi IL-10-producing macrophages was significantly increased. Dead MSCs contributed to the recruitment of macrophages to the liver by increasing PS and inducing a switch to an anti-inflammatory M2 phenotype [121]^[26]. After syngeneic transplantation of the CM-Dil-stained BM-MSCs into the fibrotic liver, CM-Dil⁺ signals were detected in CD11B⁺F4/80⁺ macrophages after 48 h, while dendritic cells and neutrophils showed minor signals. Thus, macrophage infiltration plays a major role in the clearance of infused apoptotic BM-MSCs. Moreover, the main part of the signal fell on Ly6Clo macrophages. Ly6Clo macrophages are highly restorative in fibrosis by upregulating various MMPs including MMP9, MMP12, and MMP13, which promote matrix degradation. The in vitro experiments have shown that MMP12 is the main effector in Ly6Clo macrophage-mediated resolution of fibrosis upon stimulation with BM-MSC ApoBDs. For the in vivo analysis, BM-MSCs, ApoBDs, and BM-MSCs pretreated with the caspase inhibitor Z-VAD-FMK were transplanted into fibrotic mice for the treatment of hepatic fibrosis. As expected, BM-MSC-infused livers showed low levels of hydroxyproline and α-SMA. However, treatment with only ApoBDs and BM-MSCs treated with Z-VAD-FMK showed no significant improvement in fibrosis compared to the PBS control group. These results demonstrate that the apoptosis of transplanted cells must occur in vivo in order to realize an anti-fibrotic therapeutic effect [123]^[28]. Endothelial cells (ECs) and platelets were also found among the cells capable of phagocytizing transplanted MSCs or MSC-ApoBDs and, therefore, acting as mediators between the infused cells and therapeutic effects. In a model of myocardial infarction, Liu et al. [109]^[44] found that MSCs transplanted intramyocardially into the border zone of the infarction release ApoBDs enhancing angiogenesis and improving functional recovery of the heart. Endothelial cells were identified as the main cells engulfing apoptotic MSCs in the myocardium. After intramyocardial injection, apoptotic MSCs were internalized by the PECAM1/CD31-positive ECs by 48 h. In addition, apoptotic MSCs were occasionally phagocytosed by VIM-positive cardiofibroblasts and TNNT2-positive cardiomyocytes in vivo. In vitro experiments confirmed that the main therapeutic effect of transplanted MSCs and/or apoptotic MSCs was associated with their phagocytosis by resident endothelial cells with subsequent regulation of macroautophagy/autophagy. In part, the improvement in the state of myocardial tissue could also be associated with the inhibition of the apoptosis of cardiomyocytes by apoptotic MSCs [109]^[44]. In another study [138]^[45] it was shown that ApoBDs derived from human deciduous pulp stem cells were also engulfed by ECs and increased the expression of angiogenic genes, leading to pulp revascularization and tissue regeneration in a nude mouse model of dental pulp regeneration. ApoBDs carried mitochondrial Tu translation elongation factor, which was thus transported to ECs and regulated angiogenic activation via the transcription factor EB-autophagy pathway. In an in situ beagle model of dental pulp regeneration, ApoBDs recruited endogenous ECs and facilitated the formation of dental pulp-like tissue rich in blood vessels [138]^[45]. hBM-MSCs ApoBDs corrected hemostasis in a hemophilia A model in factor VIII knockout mice through the upregulation of platelet activity following intraperitoneal injection. ApoBDs were already present on the surface of platelets 1 h after the injection. This interaction and subsequent platelet activation were mediated through the binding of Fas exposed on the surface of ApoBDs and FasL on the platelets. The ApoBD injection significantly enhanced the generation of platelet-derived microparticles. Although ApoBDs induced higher numbers of activated CD62P⁺ platelets and CD62P⁺ platelet-derived microparticles, they failed to elevate more TF⁺ platelets and TF⁺ platelet-derived microparticles. These data suggest that ApoBDs rebalance coagulation through upregulation of platelet activity without altering other hemostatic factors [112]^[46]. De Witte et al. [119]^[24] intravenously infused the PKH26-labeled human umbilical cord MSCs into the healthy BALB/c mice and found that the transplanted cells were phagocytosed by innate immune cells after 24 h. In the lungs, MSCs were engulfed mostly by the SSC⁺⁺CD11b⁺⁺ neutrophils and CX3CR⁺⁺CD11b⁺⁺ monocytes originating from the blood, and to a lesser extent by the CD68⁺CD11b⁺ macrophages. In the peripheral blood, the CX3CR⁺⁺CD11b⁺⁺ monocytes made up the main part of phagocytic cells, and neutrophils accounted for a very small part. In the liver, PKH26⁺ cells were predominantly engulfed by the CLEC4F⁺CD11b⁺ Kupffer cells, and a small percentage—by the CLEC4F^-CD11b⁺⁺ monocyte-derived macrophages and neutrophils [119]^[24]. In vitro experiments confirmed that human classical CD141⁺/CD16^- monocytes were able to phagocytize human umbilical cord MSCs, which led to their polarization towards the non-classical CD141⁺CD16⁺CD206⁺ phenotype and the expression of programmed death ligand-1 and IL-10. Monocytes primed with human UC-MSCs induced the formation of Foxp3⁺ regulatory T cells in mixed lymphocytic reactions. These results demonstrate that injected MSCs were rapidly ‘efferocytized’ by monocytes, which subsequently migrated from the lungs to other areas of the body. Phagocytosis of UC-MSCs induced phenotypic and functional changes in monocytes and subsequent modulation of the cells of the adaptive immune system [119]^[24]. The following conclusions can be drawn from the above examples: (1) monocytes/macrophages are the main cells that efferocytosis apoptotic MSCs or MSC-derived ApoBDs upon systemic administration and subsequently play a decisive role in mediating, distributing, and transmitting the immunomodulatory effects of MSCs; (2) when administered locally, apoptotic MSCs can undergo phagocytosis not only by monocytes/macrophages but also by other types of resident cells, including, for example, endothelial cells or platelets, leading to their activation and stimulation of their functions. Thus, the main mechanism of action of apoptotic MSCs or MSC-derived ApoBDs is considered to be mainly their immunomodulatory effect, due to the presence of their obvious phagocytosis by immune cells after transplantation. However, as shown in many studies, apoptotic cells and/or ApoBDs that circulate in the bloodstream, in addition to immunomodulation, are able to exert a wide variety of effects and perform various functions. For example, ApoBDs generated by tumor cells and entering the circulation have pleiotropic effects, including enhancement or suppression of antitumor immunity, promotion of metastasis, or increased procoagulant activity [140]^[47]. Many studies have also shown that apoptotic cells stimulate the proliferation of neighboring cells, triggering the so-called apoptosis-induced compensatory proliferation [141]^[48], as well as stem and progenitor cell proliferation, enhancing organ and/or tissue regeneration [142]^[49]. ApoBDs derived from cardiomyocytes and fibroblasts had different effects on myocardial regeneration in a model of doxorubicin-induced cardiomyopathy. Cardiomyocyte-derived ApoBDs stimulated the development of cardiomyocyte progenitor cells and myocardium regeneration, while fibroblast-derived ApoBDs stimulated endothelial progenitors and had no therapeutic effects [143]^[50]. The literature describes many examples of various effects, functions, and mechanisms of action of apoptotic cells and/or ApoBDs derived from various cell types [144]^[51]. Some of the mechanisms already described with regard to non-MSC-derived ApoBDs may also be mediated by apoptotic MSCs or MSC-derived ApoBDs. However, these potential mechanisms require further study.